The Challenge In developing new products and services, companies frequently have to choose from large sets of features or attributes. Deciding which will be the most popular--and profitable--is a challenge that can determine the success or failure of a new offering.
Recently, a financial services client approached us with a battery of features for a new Internet tool. The company had a list of forty-four different attributes for consideration, divided amongst five key areas: ten account management attributes, eight bill payment attributes, twelve investment attributes, six retirement planning attributes, and eight tax planning attributes. The client needed to know which features to include in its basic offering and what price to charge for the package. The problem required a research approach that effectively sorted the large battery of attributes and alleviated the need to test and retest the concepts.
There are a variety of methods commonly used to sort lists into the most promising candidates. Common techniques include management screening, focus groups, and sequential series of concept testing. However, these methods would not adequately address the client's needs in this case. Management screening often leaves highly desirable consumer attributes on the cutting room floor and often brings pet attributes to the concept. Focus groups, while attractive, do not always provide a robust, representative sample. Sequential monadic tests often suffer from potential sample error. For example, if the ultimate goal is a 40% top two box response rate for purchase intent and the true mean is 38%, then if we test it often enough we will get 40% by chance. Further, with the large number of features the client proposed, there were too many combinations to test using monadic or sequential monadic methods. Even traditional choice techniques (conjoint or discrete choice) often have trouble with so many attribute levels.
The Approach To solve the client's problem, we began with an importance sort: `How important is it that this feature be built into a product that you would purchase?' To avoid the common response that everything is important, we added a new top box, labeled `absolutely essential.' An attribute was considered absolutely essential if the consumer would not purchase the product if the attribute were missing--a deal breaker. Thus respondents sorted each feature into one of four categories: (1) absolutely essential, (2) must have, (3) nice to have, and (4) not at all necessary.
The next step was to use TURF (Total Unduplicated Reach and Frequency) analysis to determine how many features needed to be built into the basic product in order to maximize reach. The results of the TURF analysis indicated that little incremental value was gained by including more than six features. However, respondents considered some items as deal breakers, thus we needed to sort the features by the number of people that put them in the `absolutely essential' category. This stage of analysis suggested that fourteen items needed to be included in the financial services company's basic Internet tool package.
Five of the `absolutely essential' items were in the bill payment category and five in the tax planning category; clearly, the company's basic product needed to address these critical areas. Three additional features were in the account management category, but further consideration suggested a strong overlap with some of the features in the bill payment category. Do the results mean that the client should drop the investments and long term planning features? No, the results showed that while those features did not need to be in the basic product, they could provide the basis for an enhanced version of the product sold at a higher price point.
The next research phase was determining a pricing strategy. To do this, we used the van Westendorp pricing model. Respondents were given the opportunity to price basic and premium bundles. The analysis was run for each group of features. The results suggested that the optimal price for the `absolutely essential' items was between $6 and $20 per month. Adding the `must have' features added $4 to the value of the product, thus providing a range of prices for both the basic and premium bundle offers.
The Results Our approach to the client's challenge quickly and effectively identified the key attributes for their new product offer. In one stage of research we were able to sort through a large battery of items, create two distinct product offers (a basic and a premium bundle), and offer pricing thoughts for each product. This helped shorten the product development time and refine the concepts that need to be tested for the next step, giving the client a competitive advantage.